Saved in:
Bibliographic Details
Main Authors: Farias, Vivek, Gijsbrechts, Joren, Khojandi, Aryan, Peng, Tianyi, Zheng, Andrew
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2406.01939
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916615756447744
author Farias, Vivek
Gijsbrechts, Joren
Khojandi, Aryan
Peng, Tianyi
Zheng, Andrew
author_facet Farias, Vivek
Gijsbrechts, Joren
Khojandi, Aryan
Peng, Tianyi
Zheng, Andrew
contents Simulating a single trajectory of a dynamical system under some state-dependent policy is a core bottleneck in policy optimization (PO) algorithms. The many inherently serial policy evaluations that must be performed in a single simulation constitute the bulk of this bottleneck. In applying PO to supply chain optimization (SCO) problems, simulating a single sample path corresponding to one month of a supply chain can take several hours. We present an iterative algorithm to accelerate policy simulation, dubbed Picard Iteration. This scheme carefully assigns policy evaluation tasks to independent processes. Within an iteration, any given process evaluates the policy only on its assigned tasks while assuming a certain "cached" evaluation for other tasks; the cache is updated at the end of the iteration. Implemented on GPUs, this scheme admits batched evaluation of the policy across a single trajectory. We prove that the structure afforded by many SCO problems allows convergence in a small number of iterations independent of the horizon. We demonstrate practical speedups of 400x on large-scale SCO problems even with a single GPU, and also demonstrate practical efficacy in other RL environments.
format Preprint
id arxiv_https___arxiv_org_abs_2406_01939
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Speeding up Policy Simulation in Supply Chain RL
Farias, Vivek
Gijsbrechts, Joren
Khojandi, Aryan
Peng, Tianyi
Zheng, Andrew
Artificial Intelligence
Distributed, Parallel, and Cluster Computing
Machine Learning
Simulating a single trajectory of a dynamical system under some state-dependent policy is a core bottleneck in policy optimization (PO) algorithms. The many inherently serial policy evaluations that must be performed in a single simulation constitute the bulk of this bottleneck. In applying PO to supply chain optimization (SCO) problems, simulating a single sample path corresponding to one month of a supply chain can take several hours. We present an iterative algorithm to accelerate policy simulation, dubbed Picard Iteration. This scheme carefully assigns policy evaluation tasks to independent processes. Within an iteration, any given process evaluates the policy only on its assigned tasks while assuming a certain "cached" evaluation for other tasks; the cache is updated at the end of the iteration. Implemented on GPUs, this scheme admits batched evaluation of the policy across a single trajectory. We prove that the structure afforded by many SCO problems allows convergence in a small number of iterations independent of the horizon. We demonstrate practical speedups of 400x on large-scale SCO problems even with a single GPU, and also demonstrate practical efficacy in other RL environments.
title Speeding up Policy Simulation in Supply Chain RL
topic Artificial Intelligence
Distributed, Parallel, and Cluster Computing
Machine Learning
url https://arxiv.org/abs/2406.01939